Skip to content

Latest commit

 

History

History
49 lines (35 loc) · 2.8 KB

README.md

File metadata and controls

49 lines (35 loc) · 2.8 KB

Seoul-Bike-Sharing-Demand-Prediction

Superivised Machine Learning- Regression

Currently Rental bikes are introduced in many urban cities for the enhancement of mobility comfort. It is important to make the rental bike available and accessible to the public at the right time as it lessens the waiting time. Eventually, providing the city with a stable supply of rental bikes becomes a major concern. The crucial part is the prediction of bike count required at each hour for the stable supply of rental bikes. image

Attribute Information:

Data Description

The dataset contains weather information (Temperature, Humidity, Windspeed, Visibility, Dewpoint, Solar radiation, Snowfall, Rainfall), the number of bikes rented per hour and date information.

Attribute Information:

  • Date : year-month-day

  • Rented Bike count - Count of bikes rented at each hour

  • Hour - Hour of he day

  • Temperature-Temperature in Celsius

  • Humidity - %

  • Windspeed - m/s

  • Visibility - 10m

  • Dew point temperature - Celsius

  • Solar radiation - MJ/m2

  • Rainfall - mm

  • Snowfall - cm

  • Seasons - Winter, Spring, Summer, Autumn

  • Holiday - Holiday/No holiday

  • Functional Day - NoFunc(Non Functional Hours), Fun(Functional hours)

Conclusions:

  • As we have calculated MAE, MSE, RMSE, R2 score and Adjusted R2 score for each model. Based on Adjusted R2 score will decide our model performance.

  • Linear,Lasso,Ridge and ElasticNet.

  • 1)linear,Lasso,Ridge and Elastic regression models have almost similar Adjusted R2 scores(61%) on test data.(Even after using GridserachCV we have got similar results as of base models).

Decision Tree Regressor:

  • After hyperparameter tuning we got Adjusted R2 score as 84% (approx.) on test data which is quite good for us.

Random Forest:

  • On Random Forest regressor model, without hyperparameter tuning we got Adjusted R2 score as 90% (approx.) on test data. Thus our model memorised the data.So it was a overfitted model, as per our assumption

  • After hyperparameter tuning we got Adjusted R2 score as 87%(approx.) on test data which is very good for us.

Gradient Boosting Regression(Gradient Boosting Machine):

  • On Gradient Boosting regressor model, without hyperparameter tuning we got Adjusted R2 score as 86% (approx.) on test data.Our model performed well without hyperparameter tuning.

  • After hyperparameter tuning we got Adjusted R2 score as 91% (approx.) on test data,thus we improved the model performance by hyperparameter tuning.

Thus Gradient Boosting Regression(GridSearchCV) and Random forest(gridSearchCv) gives good Adjusted R2 scores. We can deploy this models.